Verified Commit 76ce0023 authored by Max R. P. Grossmann's avatar Max R. P. Grossmann
Browse files

Further clarification

parent e79721b4
......@@ -3,3 +3,19 @@
AnonPay helps to enhance privacy in online experiments. **For the first time in experimental history, AnonPay made it possible to safely collect payment details within the experiment.**
Included in this repository are slides of a recent talk on AnonPay as well as example experiments for oTree and z-Tree. In all, these materials should be sufficient to use these tools. If you have further questions, please contact us: <>
## Overview
There are 5 algorithms included in AnonPay:
- Privacy enhancement through *payment adjustment*
1. ROD: Rounding off payment amounts
2. NUN: Ensuring nonuniqueness of payment amounts
3. SPA: Enhance privacy for specially protected attributes
- Privacy enhancement through *data erasure*
4. CPY: Separate payment details from behavioral dataset
5. TDY: Redact the behavioral dataset
ROD is not recommended (see the slides, p. 21; rounding is inefficient).
It is optional to use NUN and SPA. However, everyone should use CPY and TDY (which must always be used in combination) or similar techniques. If no sophisticated programming language is available, I recommend the "payment form methodology" (see slides, p. 8). This, while trivial to implement, is beyond the scope of AnonPay.
......@@ -43,6 +43,8 @@ class Group(BaseGroup):
paym = anonpay.ROD(paym, 0.1) # the second argument here is chi
paym = anonpay.NUN(paym)
paym = anonpay.SPA(paym, attr, 0.4) # the third argument is eta
# These algorithms are OPTIONAL; they are not required to be used in conjunction
# with the AnonPay app. However, what follows is not optional:
# After the algorithms, the adjusted payments are written back:
for (p, pi_tilde) in zip(self.get_players(), paym):
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment